The long-term goal of the proposed study is to determine the nature and functional significance of structural diversity among class l antigen presenting molecules. Structural diversity in the class I peptide binding site is known to influence the way individuals respond to specific antigens by determining which antigens can be presented and by shaping the adult T cell receptor repertoire. How this is accomplished at the molecular level, however, is still being described. The studies proposed here focus on the laboratory mouse because the class I molecules of this species closely resemble the profiles of diversity and immunologic function of class I molecules in humans, and as a laboratory animal, the mouse is amenable to experimental manipulation. Recent studies illustrate the importance of amino acid diversity among class I molecules to their peptide binding and presentation properties. The hypothesis that amino acid diversity in the class I peptide binding can be generally attributed to effects on peptide binding will be examined by investigating three apparent exceptions to this rule. T cell receptors are believed to make contact with the o: helical region of class I molecules and form an interface with the exposed portions of peptides bound for presentation. How this results in the observed specificity at the molecular level will be addressed by examining the molecular basis of recognition of two structurally distinct class I molecules by the same T cell receptor. Diversity within the antigen binding site influences the development of the adult T cell repertoire. Several lines of recent evidence indicate that this occurs as a consequence of differences in peptide binding during both positive and negative selection. While the importance of peptide presentation in negative selection is generally understood, no examples of self peptides that function in positive selection have been described. A scheme using T cell receptor transgenic mice is presented for identifying a self peptide believed to function in positive selection. Furthermore, genetic variants of a class I molecule known to be important in the positive selection of several different T cell receptors will be used in a bone marrow transplant model to analyze the relationship between positive selection, peptides bound, and the antigens the selected T cell receptors are capable of responding against as part of the adult T cell repertoire.